v1.7

• Fixed a bug in the antiderivative branch of the TMS correction (apologies that that part of the code has been buggy)
• Expanded coverage of I_test.py, in part guided by the recent bugs that were found
• Multiple optimizations by Claude Opus 4.7. Most significantly, the for-loops over the layers have almost all been vectorized. Every change was personally verified

I suspect this is about as optimized as PythonicDISORT can get while keeping to only NumPy and SciPy.

Full Changelog: v1.6...v1.7

v1.6

In v1.5 it was possible to get divisions by 0 in the TMS corrections (part of the NT corrections) for very thick atmospheres. That has been fixed.

Full Changelog: v1.5...v1.6

v1.5

• Properly fixed antiderivative / integration functionality; otherwise code mostly the same as v1.3 (v1.4 has been superseded)
• Added Test I to test antiderivative / integration functionality
• Updated CITATION.cff (please cite JOSS paper as well) and CONTRIBUTING.md

Full Changelog: v1.3...v1.5 (v1.4 skipped)

v1.3

Major optimizations in almost every part of the code.

For users, the only function change is that cache_BDRF_Fourier_modes(N, mu0, BDRF_Fourier_modes) has been changed to cache_BDRF_Fourier_modes(N, BDRF_Fourier_modes, mu0=0), i.e. cache_BDRF_Fourier_modes can now cache with respect to a fixed number of streams or w.r.t. both a fixed number of streams and a fixed $\mu_0$.

What's Changed

• Bump actions/download-artifact from 4.1.9 to 5.0.0 by @dependabot[bot] in #34
• Bump actions/checkout from 4 to 5 by @dependabot[bot] in #35
• Bump actions/setup-python from 5 to 6 by @dependabot[bot] in #36
• Bump pypa/gh-action-pypi-publish from 1.12.4 to 1.13.0 by @dependabot[bot] in #37

Full Changelog: v1.2...v1.3

v1.2

• pydisort now internally enforces Kirchoff's law of thermal radiation for the isotropic internal source (its emissivity is set to 1 - omega_arr). Consequently, emissivity no longer needs to be factored into s_poly_coeffs and subroutines.generate_s_poly_coeffs has been changed accordingly.
• Streamlined code in PythonicDISORT. This should not cause any noticeable change in functionality or performance.

Full Changelog: v1.1...v1.2

v1.1

• Fixed bug wherein having a layer with no scattering inbetween two scattering layers will cause miscalculations
• Implemented zeroing of tiny single-scattering albedo $\omega$ values or tiny Legendre coefficients -- it will still be better for users to zero such values themselves
• General improvements to documentation including a new warning that using an excessive number of streams $\text{NQuad} \gtrsim 300$ will cause some quadrature angles to be extremely glancing ($\mu \sim 0$) and the computed intensity at those angles may be inaccurate

Full Changelog: v1.0.3...v1.1

v1.0.3

What's Changed

• Upgraded the BDRF_Fourier_modes argument in pydisort such that the list can now contain functions and scalars
• Upgraded subroutines.cache_BDRF_Fourier_modes and updated documentation accordingly
• Bump actions/download-artifact from 4.1.8 to 4.1.9 by @dependabot in #31
• Bump pypa/gh-action-pypi-publish from 1.12.2 to 1.12.4 by @dependabot in #30

Full Changelog: v1.0.1...v1.0.3 (v1.0.2 skipped)

v1.0.1

The restriction on the phase function Legendre coefficients Leg_coeffs_all is corrected from $[0, 1]$ to $[-1, 1]$.

Full Changelog: v1.0.0...v1.0.1

v1.0.0

v1.0.0 is finally out! We added subroutines to facilitate the modeling of longwave radiative transfer. In addition, we implemented almost all of DISORT's longwave radiative transfer tests (tests 6 and 7 in particular) and tested PythonicDISORT on the anisotropic Hapke BDRF.

We are now confident that PythonicDISORT matches DISORT in functionality.

Full Changelog: v0.9.5...v1.0.0

v0.9.5

What's Changed

• Fixed bug in computing the antiderivative of u0
• Improved PythonicDISORT.subroutines.interpolate such that it can now interpolate u0
• Changed the default of use_banded_solver_NLayers from 13 to 10 (the associated git commit message states "10 to 13" but that is a typo)
• Bump pypa/gh-action-pypi-publish from 1.10.1 to 1.10.3 by @dependabot in #22
• Bump pypa/gh-action-pypi-publish from 1.10.3 to 1.12.2 by @dependabot in #27

Full Changelog: v0.9.3...v0.9.5 (v0.9.4 skipped)